Canvas warnings for misconfigured nodes need improvement in three areas.
1. Warning UI design
The current warning output in the console is not well structured. It could use better visual hierarchy and formatting. Node references are plain links — they should be richer UI elements, similar to the node mention chips we use in the agent chat.
2. Missing warnings
Several misconfiguration scenarios do not produce warnings at all. One example:
- Action node with no inputs: If an action node is properly configured but has no incoming edges, it will never be triggered. There is currently no warning for this, but there should be — an action node without any input is effectively dead.
We should audit for other similar cases where nodes are configured but will never execute due to missing connections.
3. Placement and rendering
Warnings should be surfaced in a more visible and accessible location. The current placement can be easy to miss. Worth exploring better options for where and how warnings are rendered so users notice them before they run into issues.
Canvas warnings for misconfigured nodes need improvement in three areas.
1. Warning UI design
The current warning output in the console is not well structured. It could use better visual hierarchy and formatting. Node references are plain links — they should be richer UI elements, similar to the node mention chips we use in the agent chat.
2. Missing warnings
Several misconfiguration scenarios do not produce warnings at all. One example:
We should audit for other similar cases where nodes are configured but will never execute due to missing connections.
3. Placement and rendering
Warnings should be surfaced in a more visible and accessible location. The current placement can be easy to miss. Worth exploring better options for where and how warnings are rendered so users notice them before they run into issues.